
Energy Dissipation in Synchronous Binary Asteroids
Love number 𝑘2(Murray and Dermott,1999), or the con-
stant time lag model, in which the angle between the tidal
bulge and the line connecting the two bodies is a constant
Δ𝑡(Mignard,1979;Hut,1981). Based on the physics of
our problem setup, the secondary will librate about the syn-
chronous configuration, and thus a constant lag angle would
be inappropriate, as the lag angle should oscillate as a re-
sult of the libration. For this reason we select the constant
𝑄model, which is the same model adopted by Jacobson and
Scheeres (2011a), in which the tidal torque is defined as:
Γ=−sign(𝜔−𝜔𝑜𝑟𝑏)3
23
4𝜋𝜌 2𝐺𝑀2
𝐴𝑀2
𝐵
𝑟6𝑅
𝑘
𝑄(8)
where the body’s angular velocity is 𝜔, the orbit’s angular
rate is 𝜔𝑜𝑟𝑏,𝑅is the reference radius for the body, 𝜌is its den-
sity, and 𝑄∕𝑘is the tidal dissipation ratio. The tidal quality
factor 𝑄is related to the tidal bulge lag angle (𝑄∼ 1∕ sin 𝜖),
while the love number 𝑘2describes the level of body defor-
mation due to the tidal potential. Henceforth we drop the
subscript 2 on the Love number for simplicity. A large value
of 𝑄∕𝑘corresponds to a more stiff body that dissipates more
slowly. In reality, the tidal dissipation is far more compli-
cated than simply selecting constant values for the unknown
𝑄∕𝑘values. As pointed out by Efroimsky (2015), tidal dis-
sipation in binary asteroids, including rubble piles, may be
governed primarily by the body’s viscosity, rather than rigid-
ity. Others, including Goldreich and Sari (2009) and Nimmo
and Matsuyama (2019), argue that friction is a critical pa-
rameter. Since there is no current estimate for the viscosity
of rubble pile asteroids to the authors’ knowledge, we adopt
the friction approach. Furthermore, while many studies as-
sume the quality number 𝑄to be constant, this parameter
depends on the tidal frequency. Further complicating this re-
lationship is the fact that the tidal quality number is not a lin-
ear function of the tidal frequency, and can either increase or
decrease with the frequency (Ferraz-Mello,2013). We also
are left with the problem of the tidal lag angle oscillation.
To address this we adopt the same solution as Jacobson and
Scheeres (2011a); we will linearize the tidal torque around
the point where (𝜔−𝜔𝑜𝑟𝑏)is near zero so the torque does
not immediately switch signs as the secondary librates (see
Appendix C therein for details on this linearization). This
linearization is necessary, as the tidal bulge is a physical phe-
nomenon and requires a finite time to cross between leading
or trailing the tide-raising body.
Note that Taylor and Margot (2010) point out the simple
tidal model assumes the two bodies are widely separated,
whereas the separation between Didymos and Dimorphos
is only slightly larger than 3 primary radii. In their work,
Taylor and Margot (2010) calculate that higher order terms
in the tidal potential speed up the process of tidal evolu-
tion. However, they also find that uncertainties in the sys-
tem, particularly surrounding 𝑄∕𝑘, dominate over the higher
order tidal expansion. Thus, we continue with the simple
tidal model given the large uncertainty associated with the
bulk system density and physical properties, while keeping
in mind higher order terms in the tidal model will only in-
crease the rate of damping in the system. Thus, the error as-
sociated with this tidal model is considered to be secondary
to the considerable uncertainty on the 𝑄∕𝑘coefficient for
our purposes.
Unfortunately, given the lack of knowledge on the phys-
ical parameters of rubble piles, particularly their viscosity,
we cannot calculate an accurate value for 𝑄∕𝑘. For lack of a
better option, we surrender ourselves to the typical simplifi-
cations surrounding the factor 𝑄∕𝑘, and we turn to the work
by Nimmo and Matsuyama (2019), who derive an estimate
for a constant 𝑄∕𝑘for rubble pile binary asteroids, which
can be approximated by
𝑄
𝑘≈ 300𝑅(9)
for 𝑅in meters. This leads to a value for the primary 𝑄𝐴∕𝑘𝐴≈
1 × 105and for the secondary 𝑄𝐵∕𝑘𝐵≈ 2.5 × 104. Note this
expression for 𝑄is frequency dependent and derived for a
non-synchronous binary system. However, we again empha-
size there is large uncertainty associated with these values so
this definition serves as a first-order approximation, as the
error from uncertainties dominates over the error from the
assumptions. Furthermore, these values are consistent with
existing estimates for small bodies in the literature (Brasser,
2020;Jacobson and Scheeres,2011b;Scheirich et al.,2015,
2021). However, there is not a consensus on this linear scal-
ing. For example, Goldreich and Sari (2009) propose an in-
verse scaling of 𝑄∕𝑘with 𝑅, and even in their own work
Nimmo and Matsuyama (2019) point out a scaling with 𝑅3∕2
may be more accurate. Another consideration is if Dimor-
phos turns out to be monolithic instead of a rubble pile, its
𝑄∕𝑘value would likely be orders of magnitude higher (Gol-
dreich and Sari,2009). Hence, the large uncertainty in 𝑄∕𝑘
dominates over other errors associated with our model, and
it is futile to develop a high fidelity tidal model while lim-
ited by this unknown parameter. Given the large uncertainty
and lack of consensus around 𝑄∕𝑘, we adopt the linear scal-
ing only as nominal parameters, and subsequently investi-
gate how varying 𝑄∕𝑘for both the primary and secondary
affects the system behavior later in Section 8.
Returning to the tidal torque equation, this model is still
only defined in 2-dimensions and we wish to extend this to
a full 3-dimensional analysis, as out-of-plane rotation of the
secondary is a possibility. This can be done with only a few
corrections to the classic model. To start, we need to de-
fine a vector for the torque direction. The torque will act
to push the spin rates of the asteroids into the synchronous
equilibrium, but physically cannot act in the direction of the
position vector of the secondary relative to the primary. We
define the relative spin rate of a body:
̇
⃗
𝜙=⃗𝜔 −⃗𝜔𝑜𝑟𝑏.(10)
We can then use this spin vector as the vector along which
the torque acts, with a small correction so the torque in the
radial direction is zero:
̂
Γ=−
̇
⃗
𝜙− ( ̇
⃗
𝜙⋅̂𝑟)̂𝑟
̇
⃗
𝜙− ( ̇
⃗
𝜙⋅̂𝑟)̂𝑟
(11)
A. J. Meyer et al.: Preprint submitted to Elsevier Page 5 of 24